Incident Manager, Safeguards (AI)

Location: San Francisco, CA; New York City, NY; or Washington, DC. Hybrid policy requires staff to work from one of the offices at least 25% of the time; some roles may require more office time. After-hours and weekend incident coverage is required on a rotating basis.

Annual salary: $245,000–$285,000 USD.

Company

Anthropic develops reliable, interpretable, and steerable AI systems intended to be safe and beneficial for users and society.

What you will do

• Serve as incident commander for Safeguards incidents from declaration through closure, assessing severity and coordinating responders across Legal, Communications, Policy, Threat Intelligence, Security, Engineering, and research.
• Track facts, changes, open issues, decisions, and action items; provide regular updates to executives and partner teams.
• Coordinate legal, communications, public policy, and external notifications when required.
• Oversee the Safeguards Enforcement on-call queue, triage urgent requests and escalations, resolve or route them, and identify emerging incidents.
• Run post-incident reviews, maintain incident records and performance metrics, and track assigned actions through closure.
• Improve the incident response program by developing runbooks, standard operating procedures, automation, incident readiness, metrics, and decision rules with Safeguards engineers.

Requirements

• Experience running incidents or urgent escalations from start to finish in trust and safety, security, platform integrity, crisis management, or a comparable field, with readiness to lead high-severity incidents with support.
• Hands-on experience handling product or platform abuse or misuse, such as account compromise, fraud, or coordinated malicious activity.
• Ability to make clear decisions under time pressure, document reasoning, and recognize decisions that belong to Legal, Policy, or leadership.
• Ability to write concise, accurate updates suitable for executive action and legal or external scrutiny.
• Experience coordinating cross-functional participants who do not report to you, securing commitments, and following through.
• Experience independently improving operational processes and distinguishing which steps require human judgment versus automation.
• Willingness to perform hands-on queue work regularly and use queue activity to identify operational problems.
• Ability to participate in rotating after-hours and weekend incident coverage.
• Bachelor’s degree or an equivalent combination of education, training, and experience; relevant field of study demonstrated through coursework, training, or professional experience.

Nice to have

• Experience joining an incident response function early and helping create runbooks and review processes.
• Experience with incidents involving AI systems, enforcement actions, threat intelligence, security investigations, regulatory or external reporting, or disclosure.
• Experience operating an on-call rotation or support escalation queue, including setting and measuring response targets.
• Experience using AI tools to automate triage, summaries, or routine decisions with quality checks.
• Working knowledge of SQL or similar tools for incident and queue data analysis.
• Experience designing and facilitating tabletop exercises.
• Genuine interest in AI safety and responsible technology development.

Culture & Benefits

• Collaborative work across research, engineering, policy, legal, security, communications, and business functions.
• Flexible working hours and a location-based hybrid office policy.
• Competitive compensation, optional equity donation matching, generous vacation, and parental leave.
• Office space designed for collaboration.
• Exposure to explicit, violent, sexual, or psychologically disturbing content may be part of the role.

Hiring process

• Anthropic provides guidance on candidates’ use of AI during the application process.
• Visa sponsorship may be available; sponsorship is evaluated by role and candidate, with immigration-lawyer support.